Can I run KAT-Dev-72B-Exp on a Radeon RX 6500 XT?
Not at these settings. No indexed quantization of KAT-Dev-72B-Exp fits Radeon RX 6500 XT at any context we compute, with q8_0 KV. The smallest shipped quantization is 22.11 GiB in weights alone, against 3.72 GiB usable. CPU offload can still run it, slowly.
Every quantization at every context
| Quant | Weights● | 4K◐ | 8K◐ | 16K◐ | 32K◐ | 64K◐ | 128K◐ |
|---|---|---|---|---|---|---|---|
| Q8_0 | 71.96 GiB | 73.7 | 74.3 | 75.6 | 78.3 | 83.6 | 94.2 |
| Q6_K | 59.93 GiB | 61.6 | 62.3 | 63.6 | 66.3 | 71.6 | 82.2 |
| Q5_K_M | 50.71 GiB | 52.4 | 53.1 | 54.4 | 57.1 | 62.4 | 73.0 |
| Q5_K_S | 47.85 GiB | 49.5 | 50.2 | 51.5 | 54.2 | 59.5 | 70.1 |
| Q4_K_L | 45.02 GiB | 46.7 | 47.4 | 48.7 | 51.4 | 56.7 | 67.3 |
| Q4_K_M | 44.16 GiB | 45.9 | 46.5 | 47.8 | 50.5 | 55.8 | 66.4 |
| Q4_1 | 42.56 GiB | 44.3 | 44.9 | 46.2 | 48.9 | 54.2 | 64.8 |
| Q4_K_S | 40.88 GiB | 42.6 | 43.2 | 44.6 | 47.2 | 52.5 | 63.2 |
| Q4_0 | 38.54 GiB | 40.2 | 40.9 | 42.2 | 44.9 | 50.2 | 60.8 |
| IQ4_NL | 38.48 GiB | 40.2 | 40.8 | 42.2 | 44.8 | 50.1 | 60.8 |
| IQ4_XS | 36.98 GiB | 38.7 | 39.3 | 40.7 | 43.3 | 48.6 | 59.3 |
| Q3_K_L | 36.79 GiB | 38.5 | 39.2 | 40.5 | 43.1 | 48.4 | 59.1 |
| Q3_K_M | 35.11 GiB | 36.8 | 37.5 | 38.8 | 41.5 | 46.8 | 57.4 |
| IQ3_M | 33.07 GiB | 34.8 | 35.4 | 36.8 | 39.4 | 44.7 | 55.3 |
| Q3_K_S | 32.12 GiB | 33.8 | 34.5 | 35.8 | 38.5 | 43.8 | 54.4 |
| IQ3_XS | 30.59 GiB | 32.3 | 32.9 | 34.3 | 36.9 | 42.2 | 52.9 |
| IQ3_XXS | 29.66 GiB | 31.4 | 32.0 | 33.3 | 36.0 | 41.3 | 51.9 |
| Q2_K_L | 28.90 GiB | 30.6 | 31.3 | 32.6 | 35.2 | 40.6 | 51.2 |
| Q2_K | 27.76 GiB | 29.5 | 30.1 | 31.5 | 34.1 | 39.4 | 50.0 |
| IQ2_M | 27.32 GiB | 29.0 | 29.7 | 31.0 | 33.7 | 39.0 | 49.6 |
| IQ2_S | 26.02 GiB | 27.7 | 28.4 | 29.7 | 32.4 | 37.7 | 48.3 |
| IQ2_XS | 25.20 GiB | 26.9 | 27.6 | 28.9 | 31.5 | 36.9 | 47.5 |
| IQ2_XXS | 23.74 GiB | 25.4 | 26.1 | 27.4 | 30.1 | 35.4 | 46.0 |
| IQ1_M | 22.11 GiB | 23.8 | 24.5 | 25.8 | 28.5 | 33.8 | 44.4 |
Figures are GiB of total memory: weights plus KV cache plus compute buffer and backend overhead. Weights and KV are near-exact; the overhead term is modeled. Hover any cell for the breakdown.
Why other calculators disagree
A parameters × bits ÷ 8 estimate ignores two things that dominate at long context. First, the weights themselves are not the nominal rate — quantizations are mixtures, so the real file is consistently larger than the label implies. Second, most of this model's layers cache only a 131,072-token window rather than the full context.